CORROSION BEHAVIOUR
OF DISC SPRINGS
TABLE OF CONTENTS
Introduction
Disc spring variants, corrosion media and test methods
Materials
Coatings
Mechanical zinc plating
Dacromet coating
Geomet coating
Delta Tone + Delta Seal coating
Chemical nickel plating
Corrosion media
Corrosion testing procedures
Immersion experiments
Results of the immersion experiments
4
4
5
5
5
6
6
6
7
8
8
9
VDA alternating corrosion test
Results of the VDA alternation test and discussion
Summary of VDA alternation test
9
10
13
Stress corrosion crack experiments
Execution of the stress corrosion crack experiments
Results of stress corrosion crack experiments and discussion
Corrosion behaviour of 6 x 1 stacked disc spring column
Stress corrosion crack behaviour in seawater atmosphere
Stress corrosion crack behaviour of the disc spring stack in a 40 % MgCl2 solution
Stress corrosion crack behaviour of the disc spring stack in a 3 % NaCl solution
Stress corrosion crack behaviour of disc spring stack in 0.1 M citric acid
Corrosion behaviour of 4 x 2 stacked disc spring columns
Summary of the stress crack corrosion experiments
13
13
14
14
17
19
22
23
27
28
VDA alternation test with mechanical load
Results of the VDA alternation test with mechanical stress and discussion
Summary of VDA alternation test with mechanical load
29
29
30
Fatigue crack corrosion experiments
Execution of the fatigue crack corrosion experiments
Results of the fatigue crack corrosion experiments
Summary of the fatigue crack corrosion experiments
32
32
32
37
Bibliography
2
3
39
INTRODUCTION
When advice is offered on the use of disc springs, questions about environmental
conditions with corrosive attack come up regularly, but in many cases, they can only
be addressed very speculatively. Therefore, especially when durability must be guaranteed over a long period of time, it makes sense to proceed with a field test.
When specifying a disc spring under corrosive conditions, the load spectrum must be
analysed precisely: corrosion medium, temperature, static or dynamic stress, friction
and wear on contact sites, and the way they change over time. When selecting a material, in addition to the ideal material properties, the actual existing properties of the
starting material play a large role, and then become the decisive selection criteria.
Other material concepts, such as the coating of materials that are not resistant to
corrosion, may then be of interest. In the transition to such alternative solutions, the
differences that emerge in a direct comparison are then of interest.
This article describes the results from a research proposal conducted as AVIF Project
Number A 210, ‘Investigations of the corrosion behaviour of disc springs and disc
spring stacks’, advised by the Spring Association (VDFI), Subcommittee on Disc
Springs, via the Technical University of Darmstadt, Institute for Materials.
Systematic results on the behaviour and durability of disc springs under applicationrelated corrosive conditions were to be determined.
3
DISC SPRING VARIANTS,
CORROSION MEDIA
AND TEST METHODS
Table 1: Disc spring
variants studied
Table 1 lists the disc spring variants that were
studied. In addition to the conventional grades
of stainless spring steel in various production variants, 51CrV4 steel was also studied with various
coatings.
No.
Material
Dimensions
Production
01
02
03
04
05
06
07
08
09
10
11
12
13
14
1.4310
1.4310
1.4310
1.4568
1.4568
1.4568
1.8159
1.8159
1.8159
1.8159
1.8159
1.8159
1.8159
1.8159
63 x 31 x 1,9 x 4,15 (C)
63 x 31 x 1,9 x 4,15 (C)
80 x 41 x 3 x 5,3 (B)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
63 x 31 x 1,8 x 4,15 (C)
Punched (S) + turned (D)
Punched (S) + turned (D) + shot-peened (K)
Punched (S) + turned (D)
Punched (S) + turned (D)
Punched (S) + turned (D) + shot-peened (K)
Punched (S) + turned (D) + Kolsterised
Mechanically zinc-plated + yellow chromatised + sealed
Mechanically zinc-plated + transp. chromatised + sealed
Coated with Dacromet 500
Coated with Geomet
Delta Tone (2x) + Delta Seal (2x)-coated
Chemically nickel-plated
Lacquered with water-thinnable paint
Oiled
MATERIALS
51CrV4 (1.8159) is a ferritic steel suitable for the
production of spring components of every kind
and acceptable for the use of all DIN springs. The
alloy supplements allow, in the case of greater
material thicknesses, and after tempering, the
formation of a uniform structure over the entire
cross-section. In addition, the alloy components
had a positive effect on the relaxation behaviour.
X10CrNi18-8 (1.4310) and X7CrNiAl17-7
(1.4568) are, according to DIN EN 10 151, rustresistant spring steels that are distinguished by
their special resistance against chemically corrosive substances. They acquire their spring abilities
through work hardening and/or heat treatment.
X10CrNi18-8 (1.4310) achieves its strength only
via cold forming and is therefore generally used
only up to a thickness of 2 to 2.5 mm. Depending
on cold forming grade, the cold forming begins to
decrease significantly at about 100°C. These materials should therefore not be used at higher temperatures.
4
The material X7CrNiAl17-7 (1.4568), up to a
thickness of 2.5 mm (at larger quantities up to
3.0 mm), is subjected to a simple hardening (annealing at 480°C) in addition to coldforming,
which enables high thermal stability up to 350°C.
The increase in stability achieved via annealing
has the advantage that for the same final strength,
less cold sforming is required than for X10CrNi18-8
(1.4310). This has a positive effect on the corrosion behaviour. This material state was used in the
present study.
The material X7CrNiAl17-7 (1.4568) is processed
in thicknesses > 2.5 mm (3 mm) in the soft solution-annealed state. The necessary strength is
then achieved through a double annealing procedure (structure tempering). Since the first annealing must occur at a temperature of 760°C,
chrome carbide is precipitated selectively at the
grain boundaries. The corrosion resistance of this
material state is thus substantially minimised.
Springs in the structure-annealed state should
only be used when hot strenght is required.
COATINGS
The coatings for the disc spring variants from
1.8159 are listed in the following table. The selected types are oriented toward actual usage and
have properties proven by experience. The coatings
were applied without phosphating.
Coating
Layer thickness
Mechanically zinc-plated + yellow chromatised + sealed
Mechanically zinc-plated + transp. chromatised + sealed
Coated with Dacromet 500
Coated with Geomet
Delta Tone (2x) + Delta Seal (2x)-coated
Chemically nickel-plated
Lacquered with water-thinnable paint
Q
Q
Q
Q
Q
Q
Q
MECHANICAL ZINC PLATING
The most frequently used zinc plating process for
disc springs is pellet plating. In this type of zinc
plating, the parts are provided with a thin copper
layer after careful cleaning in a dipping process
(currentless), then the parts are stirred in a drum
together with zinc powder and glass pellets of different sizes while a promoter is added. The treatment is terminated after a specified time, by
which 95–98 % of the added zinc is plated onto
the disc springs. Finally, the parts are chromatised
in a chromate solution. The effectiveness of chromate coating decreases again at temperatures
8 µm
8 µm
5 µm
8 µm
14+4 µm
25 µm
15 µm
Table 2: Coatings for
the disc springs from
1.8159 and layer
thicknesses
over 60°C. When the process is applied skilfully,
hydrogen-induced embrittlement of the piece
being worked is negligible.
DACROMET COATING
Dacromet is an inorganic, metallic, silvery-grey
coating of zinc and aluminium layers in a chromate compound (which contains CrVI). The lamellar
structure can reduce the advance of corrosion perpendicular to the surface through horizontal deflection (diagram below).
water (vapour)
pigment- and
filler particles
vehicle (binder)
lamellar
round, angular
metal surface
Figure 1:
Diagram of diffusion
of water in pigmented
lacquer [1]
5
The components are treated as single of bulk
goods and the coating is then burned in at temperatures < 280°C. Dacromet-coated springs
display very good resistance in the salt spray fog
test. In the conventional process, i.e., without pretreatment with an etchant, hydrogen embrittlement
is completely precluded.
The springs are dipped in a metallically pure state
into an aqueous dispersion of zinc flakes, chromic
acid, and various organic components. The layer is
burned in at high temperatures. This process can
be repeated multiple times, which can create layer
thicknesses of about 6, 8, or 12 µm for single,
double, or triple applications, respectively. After a
double application, the corrosion resistance in the
salt spray test according to DIN 50 021 SS
amounts to more than 240 h.
■ Resistance in salt spray test
according to DIN 50 021:
1000 h
■ Resistance in Kesternich test
according to DIN 50 018:
15 cycles
■ Resistance in condensation
water test according to DIN 50 017: 1000 h
The so-called »base coat« of Delta Tone, which
contains Zn and Al, can be applied like paint in
layer thicknesses between 5 und 12 µm. In the
hardened state, organic components are no longer
present in the coating, capable of isolating the
metallic pigments.
The organic »top coat« of Delta Seal works to isolate and protect against contact corrosion, and is
resistant to acids and alkalis. The frictional resistance can be reduced by application of solid lubricants.
GEOMET COATING
CHEMICAL NICKEL PLATING
As an alternative to the Dacromet product line,
which contains chrome (VI), the Geomet coating
has been developed. The goal was to achieve the
same corrosion-resistant properties as for the
Dacromet coating, which contains chrome (VI).
Analogous to the Dacromet and Delta Tone
coating, this is a coating that is not electrolytically
applied, consisting of zinc and aluminium layers
and a mineral seal that can also be provided with
an organic coating. There is no danger of hydrogen
embrittlement.
DELTA TONE + DELTA SEAL COATING
Similar to the Dacromet coating, this is an inorganic metallic pigmented coating material that,
thanks to feasible cathode protection of steel, has
been developed as an alternative to galvanic or
mechanical zinc coating. After burning in the solvent-containing coating, applied like paint (the
metallic particles are cross-linked with an inorganic binding agent at 200°C for 15 min), silvery-metallic coatings are produced.
For a large number of applications, Delta Tone
with or without organic sealing achieves extremely
good corrosion protection. There is no danger of
hydrogen embrittlement.
6
Nickel layers are generally used for precisely defined applications to protect against corrosion or
wear, or for cosmetic reasons. In this context, chemical nickel plating is used for disc springs. In this
process, the coating material is created as nickelphosphorus alloys. The level of phosphorus content affects the behaviour of the coating. The
best corrosion resistance and best ductility are
achieved at 10–13 % phosphorus. With decreasing
phosphorus content, the resistance to wear increases and the resistance to corrosion decreases.
Since hydrogen is produced as a side reaction
during the deposition process, hydrogen embrittlement cannot be completely precluded.
The following table gives an overview of the conventional coating process and its resistance in the
salt spray fog test according to DIN 50 021.
Process
Burnishing
Phosphatising
Phosphatising
Galvanic zinc-plating
Galvanic zinc-plating
Galv. zinc-plating + yellow chromat.
Galv. zinc-plating + yellow chromat.
Mech. zinc-plating
Mech. zinc-plating + yellow chromat.
Delta Seal
Delta Tone
Chemical nickel-plating
Dacromet 500-A
Dacromet 500-B
Layer
composition
Metal oxide + oil
Zinc phosphate + oil
Zinc phosphate + wax
Zinc
Zinc
Zinc + chromium
Zinc + chromium
Zinc
Zinc
Zinc ph. + org. layer + oil
Zinc ph. + zinc powder coat.
Nickel
Chromat. zinc lamellae
Chromat. zinc lamellae
CORROSION MEDIA
The environmental media shown in Table 4 were
used in the study. Six media with different character were selected to form conditions close to
those encountered in practice, found frequently in
experiments.
Saturated artificial seawater atmosphere simulates environmental conditions in a marine environment. Disc springs often serve for pretensioning
bridges or sluice gates in the vicinity of the sea.
For the studies, artificial seawater according to
DIN 50905, which has a standard salt content of
3.5 % and a pH value in the range from 7.8 to
8.2, was thus used.
Media
Saturated sea water atmosphere
40%ige MgCl2 solution (magnesium chloride)
3%ige NaCl solution (sodium chloride)
0,1 N NaOH (sodium hydroxide)
0,1 M citric acid (C6H8O7
Deionised water
Layer
Resistance in salt spray fog test
thickness
according to DIN 50 021
[µm]
[h] 200
400
600
800
0,5 – 1,5
Standard protection
10 – 15
ca. 10 – ca. 40
Q 8
Q 12
Q 8
Q 12
Q 12
Q 12
10 –15
10 –15
ca. 25
Q5
Q8
The 3% NaCl solution represents, for instance,
work conditions in car industry. In that context,
disc springs come into contact with chloride-containing solution through the use of road salt in
winter. In order, in addition, to study the effect of
chloride concentration on the corrosion behaviour
on disc springs, the 40 % MgCl2 solution was also
included in the study programme, since MgCl2 salt
is easier to dissolve at room temperature.
1000
Table 3:
Relative estimate of
corrosion protection
and its resistance in
the salt spray fog test
[2] achievable for various processes
Further applications for disc springs can also be
found in the food industry. For machines and
plants in the food industry, corrosion attacks
come not only from possible aggressiveness of the
material being processed, but often also through
intensive cleaning of the devices with acidic or
pH
7,80
5,02
5,29
12,90
2,09
7,40
Conductivity [mS/cm]
52,70
140,60
42,40
19,77
2,90
0,002
Table 4:
Environmental media
for the studies
7
alkaline solutions. The 0.1 normal NaOH and 0.1
molar citric acid are to be considered as representative of alkaline and acidic media.
Since the immersion experiments are not normalised, the standardised VDA alternation test [3]
was incorporated as an addition to the experimental programme.
Deionised water is desalinated water with low
conductivity and was used as a neutral reference
medium.
IMMERSION EXPERIMENTS
The immersion experiments show the corrosion
behaviour of the disc spring variants under simple
corrosive attack. The visual manifestations of corrosion on the disc spring are observed over the
course of the experiment (4 weeks) and a first
qualitative impression of the corrosion resistance
of the disc springs in the various media is obtained.
CORROSION TESTING PROCEDURES
In the research proposal, the following study procedures were used and several results are shown
here:
■
■
■
■
■
■
■
■
Immersion experiments without stress
VDA alternating test
Electrochemical experiments *
Stress corrosion crack experiments
with constant stress
Strain-induced stress corrosion crack
experiments
Fatigue crack corrosion experiments
VDA alternating tests with mechanical stress
Accompanying experiments (metallographic
experiment, raster electron microscopic experiment, residual stress measurement)*
The experiments were carried out at room temperature and without introduction of air into the
medium. The test specimens were hung such that
they extended halfway into the medium and were
not stressed mechanically or electrically. Only the
visual manifestations of corrosion were observed,
since in practice, the disc springs mostly break
due to stress corrosion cracking or corrosion fatigue, and local corrosion is more important. Thus,
no evaluation was made of the loss of weight.
The experiments indicated with * are not cited in
this report.
Tellerfedervarianten
Table 5: Summary of
results of immersion
experiments
✩ – good
★ – moderate
✪ – poor
● – very poor
8
1.4310/C/S + D
1.4310/C/S + D + K
1.4310/B/S + D
1.4568/C/S + D
1.4568/C/S + D + K
1.4568/C/S + D + kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
40 % MgCl2
3 % NaCl
0,1 n NaOH
0,1 m acid
★
★
★
★
✩
✩
✩
✩
★
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
★
✩
✩
✩
✩
✩
✩
✩
✪
●
✪
✪
✩
✩
✩
✩
✩
★
✩
✩
★
✪
✪
✩
✩
●
●
●
●
●
●
✪
✪
✪
●
The steel springs that have merely been oiled only
exhibit short-term corrosion protection and do
poorly.
In 3 % NaCl solution, the springs of stainless steel
behave better than those in 40 % MgCl2 solution,
but the coated springs perform worse, in part. The
springs coated with Zn behave better in the solution with high Cl content due to the formation of a
layer of Simonkolleite (Zn5[(OH)8Cl0]·H2O).
Figure 2:
Experimental setup for immersion experiments
RESULTS OF THE
IMMERSION EXPERIMENTS
None of the studied disc spring variants corroded
in 0.1 N NaOH, since a thick protective layer
formed. The water-dilutable paint dissolved in
0.1 N NaOH.
For disc spring applications in acids, stainless
steels should be used, since they showed good resistance here. The Zn coatings react with the acid
and dissolve at different speeds. Due to the pores
in the layer, the chemical nickel-plating also shows
poor effectiveness.
To provide an overview of the results, they have
been evaluated according to the following scheme
and listed in Table 5:
Good:
no recognisable visual manifestation
of corrosion
Moderate: low manifestation of corrosion (as
dots)
Poor:
covered with thin layer of corrosion
product (as blotches).
Very poor: covered with thick layer of corrosion
product (overall an area).
Since within the immersion experiment the disc
springs are only attacked corrosively (without mechanical or electrical stress), the results of the immersion experiments form a basis for evaluating
the corrosion behaviour of the disc springs under
complex stress.
Stainless steels 1.4310 and 1.4568 corrode in
the 40 % MgCl2 solution. Shot-peening improves
the behaviour somewhat for 1.4310, but for
1.4568, it degrades it. Kolsterising also has a
negative impact for 1.4568.
The coated disc springs have better corrosion resistance in 40 % MgCl2 solution than do springs
from stainless steels, with the exception of chemically nickel-plated variants. The chemically nickelplated variant is prone to blemishes in the layer in
the form of pores.
VDA ALTERNATING
CORROSION TEST
The VDA alternating corrosion test is a recognised
and frequently used process for studying corrosion, especially in the automobile industry. Since
the immersion experiments are not normalised,
the standardised VDA alternation test [3] was also
incorporated into the experimental programme as
a recognised and frequently used process for studying corrosion, especially in the automobile industry.
One cycle of the VDA alternation test according to
VDA test sheet 612-415 consists of [4]
■ 24 hour salt spray fog test according to DIN
50021 at 35°C
■ 96 h condensed water test according to DIN
50017 KFW at 40°C
■ 48 h standard atmosphere according to DIN
50014
In the salt spray chamber, the individual disc
springs were simply hung from rods. The test lasted for 4 VDA cycles (four weeks). The important
evaluation criterion here is the intensity of visual
manifestation of corrosion.
9
Fig. 3 shows all 14 tested disc spring variants
after 4 VDA alternation test cycles. No corrosion
can be seen in the shot-peened disc spring composed of 1.4310; on the other two variants, a low
number of brownish spots can be found, similar to
the non-shot-peened disc spring variant compo-
sed of 1.4568. The shot-peened variant composed of 1.4568 behaves somewhat worse, but the
difference is very low. Thus, after the four-week
regular experimental time span, these five variants
were left for a further 9 weeks in the salt spray
chamber. The Kolsterised variant composed of
1.4568 displays the worst resistance to corrosion
among the disc spring variants composed of the
various stainless steels. The spring is almost fully
covered with a large number of deep rust-brown
spots.
1.4310 / C / S + D
1.4310 / C / S + D + K
1.4310 / B / S + D
1.4568 / C / S + D
1.4568 / C / S + D + K
1.4568 / C / S + D + Kolsterised
RESULTS OF
THE VDA ALTERNATION TEST
AND DISCUSSION
Figure 3:
Visual manifestation of
corrosion on the disc
springs after 4 cycles
of the VDA test
10
Figure 3:
Continuance
Mechan. zinc-plated + yellow chrom.
Mechan. zinc-plated + transp. chrom.
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
Water-dilutable painted
Oiled
11
1.4310 / C / S + D
1.4310 / C / S + D + K
1.4568 / C / S + D
1.4568 / C / S + D + K
Figure 4:
Visual manifestation of
corrosion on the disc
springs after 13 cycles
of the VDA test
Disc spring variants
Table 6:
Comparison of results
from the immersion
experiments and VDA
alternation tests [6]
✩ – good
★ – moderate
✪ – poor
● – very poor
+2: 2 levels better
–2: 2 levels worse,
etc.
12
1.4310/C/S + D
1.4310/C/S + D + K
1.4310/B/S + D
1.4568/C/S + D
1.4568/C/S + D + K
1.4568/C/S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
VDA
alternation test
3 % NaCl
40 % MgCl2
✩
✩
✩
★
0
0
0
-1
-1
-1
-1
-1
0
0
0
0
-2
0
+1
+1
+1
0
0
0
-3
-2
0
0
-1
0
-2
0
✪
●
●
✪
✩
✩
★
✪
✪
●
The Dacromet- and Geomet-coated steel disc
springs do not corrode like the shot-peened disc
spring composed of 1.4310. The Delta Tone +
Delta Seal coated disc spring was attacked by
the corrosion medium, forming a large number
of white spots, meaning that the base material
of 1.8158 did not corrode. The mechanically zincplated disc spring variants behave even worse. The
yellow chromatised variant was covered with a
large number of deep rust-brown spots, and the
transparent chromatised variant by a small number of rust-brown spots and a large number of
white spots. The problem with blemishes in the
coating on the edges appears in the chemically
nickel-plated and the spring with water-thinnable
paint. A moderate number of rust-brown spots indicate that the base material has corroded. The
disc spring variant that has only been oiled behaves the worst and displays heavy surface corrosion.
In Table 6, the results from the VDA alternation
test and from the immersion experiments in 40 %
MgCl2 solution and in 3 % NaCl solution are compared with each other. The three experiments are
similar in that they involve a chloride-containing
solution and the presence of oxygen. They differ in
concentration and experimental temperature. The
high concentration of chloride can foster corrosion
by destroying the protective layer formed in air,
such as for the disc spring variants composed of
stainless steels, or inhibit corrosion by forming the
Simonkolleite layer, as for the disc spring variants
coated with zinc.
A high oxygen concentration has the same effect:
for the disc spring variants composed of stainless
steels, it prevents progressive corrosion by forming
an oxide layer, while at the same time, it accelerates the rate of corrosion via participation in reaction [5]. Thus it is recommended that each
metal/medium/stress combination be regarded
as its own system and treated separately. During
the VDA alternation test, the hung disc springs receive significantly more oxygen than during the immersion experiment, but less chloride, since according to VDA test sheet 612-415, they have direct contact with the salt spray fog for only 24
hours per cycle (7 days).
Figure 4 displays photographs of stainless steel
disc springs after 13 cycles of the VDA alternation
test. The disc spring composed of 1.4568, turned
and especially shot-peened, is somewhat worse
than the variants composed of 1.4310.
SUMMARY OF VDA ALTERNATION TEST
Despite the large supply of oxygen, the disc
springs composed of 1.4310 and the Dacromet
and Geomet-coated disc springs display very good
resistance to corrosion. The disc springs composed of 1.4568 are somewhat worse than those
composed of 1.4310. The Kolsterised, the mechanically zinc-plated, and the oiled-only disc springs
behave significantly worse in the VDA alternation
test.
STRESS CORROSION
CRACK EXPERIMENTS
In the stress corrosion crack experiments, the corrosion effects are studied in combination with static mechanical stress. The disc springs are first
placed under tension as 6 x 1 or 4 x 2 disc spring
stacks and then immersed in the corrosion medium. Newly designed test equipment for tension
ensures that the disc springs do not undergo contact corrosion. Like the immersion experiments,
the stress corrosion crack experiments are not
standardised, so the VDA alternation test is also
performed with mechanical stress. In addition,
strain-induced stress corrosion crack experiments
are carried out. In these experiments, the experimental temperature of the disc spring stacks
under tension is changed slowly from 40 °C to 80
°C and back in order to test the behaviour induced by temperature oscillations.
EXECUTION OF THE STRESS
CORROSION CRACK EXPERIMENTS
The preparation of the stress crack corrosion experiments runs according to the following steps
(Figure 5): Cleaning, applying tension as 6 x 1 or
4 x 2 stacks on a special device for 0.2h0, 0.4h0,
0.6h0 or 0.8h0, placing the stacks in a sealed
glass container (Figure 6) with the corrosion medium, and storing in a heated container at 40 or
80 °C. The test specimens were examined daily,
and the corrosion solution was refreshed every 2
weeks. If a fracture developed in a spring, the experiment was terminated.
13
Figure 5:
Steps in the stress
corrosion crack
experiments
RESULTS OF STRESS
CORROSION CRACK EXPERIMENTS
AND DISCUSSION
When stress corrosion cracking occurs, disc
springs principally fail due to crack formation.
Thus, the lifespan of the disc spring column until it
fractures is considered the most important characteristic for evaluation here. To analyse the effect
of the coating, the experiments were performed on
a 6 x 1 coated disc spring column as well as on a
4 x 2 coated disc spring column.
to these, there are other factors, such as the condition of the surface, possible flaws in the material, and the production state of the disc springs,
which can determine the mechanism of crack formation. The experiment results thus displayed a
relatively large spread, so the experiments were
repeated as often as possible. In each case, the
shortest lifespan was selected for evaluation. After
2500 h, the experiments in which no fracture had
occurred were terminated and treated as the limit
of resistance.
CORROSION BEHAVIOUR
OF 6 X 1 STACKED DISC SPRING
COLUMN
In the experiments it was determined that temperature and stress are the decisive factors for the
lifespan of the spring. Higher temperature means
a higher corrosion rate; the bending tension in the
disc spring is decisively responsible for crack formation.
The extent of (local) bending tension depends not
only on the pretension, but also on the notch effect from the corrosion-induced notch. In addition
Figure 6: disc spring stack under tension in
glass container
14
Disc spring variants
Seawater
MgCl2
NaCl
NaOH
Säure
1.4310 / C / S + D
1.4310 / C / S + D + K
1.4310 / B / S + D
1.4568 / C/ S + D
1.4568 / C / S + D + K
1.4568 / C / S + D + kolsterisiert
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
With water-dilutable paint
Oiled
Geölt
>2500h
>2500h
>2500h
>2500h
>2500h
284h
912h
1129h
>2500h
>2500h
620h
1057h
837h
1323h
356h
429h
1968h
140h
140h
2177h
>2500h
>2500h
>2500h
>2500h
>2500h
837h
>2500h
1752h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
738h
455h
360h
1534h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
68h
68h
891
891
526h
380h
262h
839h
Table 7 shows the shortest lifespan values for the
disc springs reached at static compression 0.8h0
(80 % compression) and experimental temperature of 80 °C.
At the end of the experiment, these two disc
spring variants show extensive corrosion of the
dissolved coating like that of the oiled-only spring
in citric acid.
All disc springs studied show very good stress corrosion crack behaviour in 0.1 N sodium hydroxide
in contrast to other media. No disc spring variant
failed in sodium hydroxide, which can be traced to
a good tolerance and rate of repassivation. Even
the disc spring variant with water-dilutable paint,
whose paint dissolved entirely after 2 days at 80
°C, as had already been seen in the immersion
experiment, is stable due to protective oxide or hydroxide protective layers that form at pH > 10.
The chemically nickel-plated disc springs failed in
all media except for sodium hydroxide similarly to
the failure for the disc springs with water-dilutable
paint. The oiled-only disc spring variant behaved
the worst during the immersion experiment but in
the stress corrosion crack experiment, it only
broke in the seawater atmosphere. In keeping with
the result for the immersion experiment, the Delta
Tone + Delta Seal coated disc spring in the 40 %
MgCl2 solution had a longer lifespan than in the
3 % NaCl solution. Most coated disc spring variants failed in the seawater atmosphere, though
not in the chloride-containing solutions. In the
deionised water, only the stainless steel disc
springs were studied, with all springs withstanding
2500 hours of the experiment without event.
The disc springs composed of stainless steels only
fractured in 40 % MgCl2 solution, with the exception of Kolsterised disc spring variants, which only
failed in seawater atmosphere.
The Dacromet- and Geomet-coated disc spring variants did not fracture in any corrosion media in
the stress corrosion crack experiment, where the
coatings dissolved after some time, since in acid
solutions of organic materials, zinc is not resistant
(Dacromet coating after about 20 hours of the experiment, Geomet coating after about 50 hours).
Table 7:
Lifespan of 6 x 1 disc
spring stacks in the
stress crack corrosion experiment at
0.8 h0 and 80°C
The experiments carried out at 80°C with failure
of at least one disc spring were performed again
at 40°C. The results are summarised in Table 8.
The number of broken disc springs at 40°C is
sharply reduced, since with the decrease in temperature, the corrosion rate falls and there is less
15
Disc spring variants
Seawater
1.4310 / C / S + D
1.4310 / C / S + D + K
1.4310 / B / S + D
1.4568 / C / S + D
1.4568 / C / S + D + K
1.4568 / C / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Tab. 8:
Table 8: Lifespan
of disc spring stack
(6 x 1 stacking) in
the stress corrosion
crack experiment at
0.8h0 and 40 °C
Figure 7:
Condition of the mechanically zinc-plated
and transparently
chromatised disc
springs after over
2500 h under 0.2h0
stress in citric acid
16
>2500h
>2500h
>2500h
>2500h
834h
>2500h
MgCl2
NaCl
NaOH
Acid
>2500h
>2500h
>2500h
>2500h
>2500h
>2500h
694h
corrosive stress. However, the chemically nickelplated disc spring variant failed in 40 % MgCl2 solution and in 3 % NaCl solution, as did the mechanically zinc-plated and painted variants in 0.1 M
citric acid. For the chemically nickel-plated disc
spring variant, it appears that the acceleration in
local corrosion of the base material at weak
points in the inert coating is more effective than
reduction of corrosion due to the low temperature.
The coatings for the mechanically zinc-plated and
painted disc springs also react very strongly with
the citric acid.
>2500h
116h
>2500h
45h
284h
>2500h
>2500h
>2500h
1917h
356h
After comparing the disc spring variants regarding
the lifespan until fracture in different corrosion
media (Tables 7 and 8), the following extracts represent the stress corrosion crack behaviour in the
different media. The visual manifestations of corrosion of the springs after termination of the experiment are considered in addition to the lifespan
until fracture, since visual manifestations of dissolution of the disc springs (without immediate crack
formation) are to be evaluated after a certain
period of time as failure (compare Figure 7).
Stress corrosion crack experiments in seawater
1.4310 / C / D
1.4310 / C / D + K
1.4310 / B / D
1.4568 / C / S + D
1.4568 / C / S + D + K
1.4568 / C / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Lifespan [h]
0
2.500
2.500
2.500
2.500
2.500
284
912
1.129
2.500
2.500
620
1.057
837
1.323
500
STRESS CORROSION
CRACK BEHAVIOUR
IN SEAWATER ATMOSPHERE
Figure 8 shows the lifespan of all disc spring variants in saturated seawater atmosphere at 0.8h0
and 80 °C. All stainless steel disc springs with the
exception of the Kolsterised ones and the Dacromet- and Geomet-coated disc springs did not
break in the seawater atmosphere. The Kolsterised
disc springs, by contrast, showed the shortest lifespan (284 hours).
The salts dissolved in the seawater consist of
about 77.8 % sodium chloride (table salt) and
about 10.9 % magnesium chloride, where chloride
ions are a main component. Chloride ions have
an especially deleterious corrosive effect, since
they make the oxide layers on stainless metals
porous [3].
The disc springs composed of stainless steel
showed a significantly better corrosion behaviour
than the Kolsterised variant (Figure 9). After the
experiment had run 2500 h, only small rust-brown
spots appeared on the outer edge (shot-peened
1.4310). No traces of corrosion could be seen on
the upper or lower faces of the disc springs.
Disc springs often corrode in the edge region to a
greater extent than on the upper and lower faces.
For coated springs, the thin or permeable coating
1.000
1.500
2.000
2.500
Figure 8:
Comparison of lifetime of all disc spring
variants in the stress
corrosion crack experiment in seawater
atmosphere (0.8 h0
and 80°C)
is not optimal protection from corrosion. For stainless steel disc spring variants, punching and turning creates deformation-induced martensite,
which is less resistant to corrosion than austenite,
in the edge region.
The disc spring variants composed of stainless
steels show good stress corrosion crack behaviour,
since the concentration of the chloride ions contained in the seawater atmosphere is not high
enough to completely destroy the protective layer.
At the same time, the protective layer is continuously repaired by the oxygen in the air. (The
springs are not immersed in the medium.) The
Kolsterised springs were the sole exception, being
covered by a thick rust-brown corrosion product
layer by the end of the experiment. The cause is
the high carbon content in the edge zone.
The Dacromet- and Geomet-coated springs looked
as new after 2500 h. These two coatings bestow
optimal protection on the slightly alloyed basic
material in the seawater atmosphere. The mechanically zinc-plated disc spring is covered with a
foamy rust-brown corrosion product layer. Due to
the low chloride concentration in the atmosphere,
no Simonkolleite (Zn5(OH)8Cl2·H2O) layer forms;
rather, loose and voluminous coatings are generated on the surface. These have no defined composition; rather, they consist of a variety of products.
In addition, the effectiveness of chromate layers
decreases at temperatures over 60°C.
17
The Delta Tone/Delta Seal-coated spring is in a
significantly worse condition than the Geometcoated spring, but better than the mechanically
zinc-plated spring. Only in the problem zone, the
edge region, are there some rust-brown dots. The
variant with the water-dilutable lacquer and the
chemically nickel-plated variant display the same
problem of a locally permeable coating. The oiledonly disc spring corroded heavily. The protective oil
bestows only short-term protection, and the high
oxygen content in the atmosphere accelerates the
corrosion.
1.4568 / C / S + D + K
1.4568 / C / S + D + kolsterisiert
Mechan. verzinkt + gelb chrom.
Delta Tone + Delta Seal
Wasserverd. lackiert
Geölt
Figure 9:
Visual manifestation of
corrosion on the disc
springs in the stress
corrosion crack experiment in seawater atmosphere (0.8h0 and
80 °C)
18
STRESS CORROSION CRACK
BEHAVIOUR OF THE DISC SPRING
STACK IN A 40 % MGCL2 SOLUTION
Figure 10 compares the lifespan of all variants in
40 % MgCl2 solution at 0.8h0 und 80 °C. Of the
coated disc springs, only the chemically nickelplated variant fractured, due to local corrosion at
the blemishes in the coating and the associated
notch effect. The low oxygen content and the high
chloride concentration in the 40 % MgCl2 solution
reduce and inhibit corrosion of the basic material
1.8159. These two features of 40 % MgCl2 solution are also the main reasons for the bad stress
corrosion crack behaviour of the disc springs composed of stainless steels, whose course of deterioration is determined by pitting corrosion.
For springs composed of the material 1.4310, series B and C (different thickness), the spring with
size B shows a longer lifespan than spring C in the
stress corrosion crack experiment, as opposed to
the electrochemical test, where spring B corrodes
more heavily. This is due to a longer phase of
crack progression, caused by the greater thickness
of the material; thus, there is a significant size-related effect.
As with the electrochemical experiment (not reported here), shot-peening is shown to have a positive effect on disc springs composed of 1.4310
and no clear effect on disc springs composed of
1.4568 with regard to stress corrosion crack behaviour. In general, shot-peening brings residual
compressive stress into the surface layer of the
springs [7]; as a result of the superposition with
external stresses, the resulting bending tension is
reduced. This only works when the surface is not
heavily corroded. The surface corrosion leads on
the one hand to an operation-induced notch effect and on the other hand to the erosion of the
layer under residual compressive stress. From the
immersion experiment, it was known that the shotpeened disc spring composed of 1.4568 corroded
relatively heavily in 40 % MgCl2 solution.
The springs composed of 1.4310 generally have
a longer lifespan than the springs composed of
1.4568. This also applies to the non-shot-peened
spring composed of 1.4310 and the shot-peened
spring composed of 1.4568. This means that the
material itself is of greater significance than its
method of preparation for determining the corrosion behaviour of the disc springs.
Almost disc spring variants made from stainless
steels are covered with a thin layer of rust-brown
corrosion product, with the exception of the Kolsterised variant. This matches the results from the
immersion experiment. For the rust- and acid-resistant heavily alloyed chrome steels and chromenickel steels, there is danger of pitting corrosion in
aqueous solutions containing halides (especially
chlorides and bromides). Halide ions are in a position to ‘break through’ the passive layers formed
Stress corrosion crack experiments in 40 % MgCl2 solution
1.4310 / C / D
356
1.4310 / C / D + K
429
1.4310 / B / D
1.968
1.4568 / C / S + D
140
1.4568 / C / S + D + K
140
1.4568 / C / S + D + Kolsterised
2.177
Mechanically zinc-plated + yellow chromatised 2.500
Mechanically zinc-plated + transparently chromatised 2.500
Dacromet-coated
2.500
Geomet-coated
2.500
Delta Tone + Delta Seal
2.500
Chemically nickel-plated
837
With water-dilutable paint
2.500
Oiled
1.752
Lifespan [h] 0
500
1.000
1.500
2.000
2.500
Figure 10:
Comparison of lifetime
of all disc spring variants in the stress corrosion crack experiment in 40 % MgCl2
solution (0.8h0 and
80 °C)
19
Figure 11:
Visual manifestation of
corrosion on the disc
springs in the stress
corrosion crack experiment in 40 % MgCl2
solution (0.8h0 and
80 °C)
1.4310 / C / S + D
1.4568 / C / S + D + kolsterisiert
Delta Tone + Delta Seal
Chemisch vernickelt
Geölt
in the air in such steels and thus to form local,
actively corrosive anodic centres. Since the passive layers conduct electrons, the undisturbed environment acts as a cathode. Rapid penetration of
the corrosion into the material is the result. The
danger of pitting in such steels rises with increasing contamination.
The notches created by pitting lead, however, to a
local elevation of tensile stress. If no intermediate
repassivation is possible, the attack will continue
20
to grow deeper, and the crack will propagate at an
increasing rate. In media containing chloride, the
mechanism of crack formation is accompanied by
pitting which then can grow sharply even if the
crack only progresses slowly. The variants depicted
in Figure 12 are thus produced according to the
rate of crack propagation.
passiv
pitting
»active«
highly active
Since the edge area corrodes most heavily and receives the most stress, the crack often appears on
the external edge of the spring. It then spreads toward the inner edge of the spring. This process is
slower in springs composed of 1.4310, since
1.4310 steel is more ductile. Susceptibility to cracking generally increases with higher bending tension and rising temperature.
(Fe3O4, Fe(OH)2) acts as a protective layer and
prevents the further progress of the corrosion. The
other reason that the Kolsterised spring did not
fracture is that here only surface corrosion occurs,
thanks to the impermeable and evenly distributed
carbon. The same is true for the oiled-only spring;
it behaves worst in the immersion experiment in
40 % MgCl2 solution.
The Kolsterised variant did not fail by fracturing,
although the spring was covered by a thick covering layer. Initially, it corrodes very heavily, but after
some time it stops, and the solution remains
clear. This thermodynamically stable covering layer
The chemically nickel-plated disc spring broke,
due to the notch effect caused by local corrosion.
On the mechanically zinc-plated, Dacrometcoated, and Geomet-coated springs, almost no
traces of corrosion can be seen. The Simonkolleite
Figure 12:
Appearance of Tran
crystalline stress corrosion crack in rust- and
acid-resistant Cr-Ni
steels in metallographic
cross-section (schematic) [8]
Stress corrosion crack behaviour of the disc spring stack in a 3 % NaCl solution
1.4310 / C / D
1.4310 / C / D + K
1.4310 / B / D
1.4568 / C / S + D
1.4568 / C / S + D + K
1.4568 / C / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Lifespan [h] 0
2.500
2.500
2.500
2.500
2.500
2.500
2.500
2.500
2.500
2.500
738
455
360
1.534
500
1.000
1.500
2.000
2.500
Figure 13:
Comparison of lifetime of all disc spring
variants in the stress
corrosion crack experiment in 3 % NaCl
solution (0.8h0 and
80 °C)
21
Figure 14:
Visual manifestation
of corrosion on the
disc springs in the
stress corrosion crack
experiment in 3 %
MgCl2 solution (0.8h0
and 80 °C)
1.4568 / C / S + D + K
1.4568 / C / S + D + Kolsterised
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
layer that is produced protects the springs against
corrosion; the other coating component, aluminium, is highly resistant to corrosion in neutral corrosion electrolytes (pH 5 to 9). The spring coated
with Delta Tone/Delta Seal is somewhat worse in
comparison; it displays some rust-brown corrosion
dots on the edges of the spring.
22
STRESS CORROSION CRACK
BEHAVIOUR OF THE DISC SPRING
STACK IN A 3 % NACL SOLUTION
Figure 13 compares the lifespan of all variants in
3 % NaCl solution at 0.8h0 und 80 °C. The stress
corrosion crack behaviour of most disc springs in
3 % NaCl solution is better than in 40 % MgCl2
solution, since the electric conductivity of 3 %
NaCl solution is lower than that of 40 % MgCl2 solution and fewer ions that would favour corrosion
on the metal surface go into solution. Here there
was no stainless steel variant that failed. In contrast, the variant with water-dilutable paint and
the variant coated with Delta Tone and Delta Seal
broke.
Figure 14 shows several disc springs after the experiment in 3 % NaCl solution. All in all, the condition of the springs is better than after the experiment in seawater atmosphere and in 40 % MgCl2
solution. Here there is a lower oxygen supply than
in seawater atmosphere, since the disc spring
stack is immersed in the solution. The lower chloride concentration has the advantage that the airpassivated layer is not heavily damaged, which is
the reason that the disc springs composed of
stainless steels are covered only by a thin
brownish layer. On the other hand, there is also
the disadvantage that the Simonkolleite layer on
the Delta Tone and Delta Seal-coated disc spring
was not fully formed via the low Cl-concentration.
White spots and rust-brown corrosion dots are the
evidence for corrosion of the coating and base
material.
For the disc spring variant with water-dilutable
paint and the Delta Tone/Delta Seal-coated variant, the NaCl solution reached the base material
through the channels in the coating. This led the
springs to break. The permeable coating is not a
sure factor for the corrosion behaviour of the disc
spring. A similar situation applies for the chemically nickel-plated disc springs due to the existing
pores in the coating, where the contact corrosion
between coating and base material leads to acceleration. The chemical nickel-plating provides no
protection here and the final failure was the breakage of the disc spring.
The oiled-only disc springs likewise failed at the
end through breakage, since the oil can only offer
short-term corrosion protection.
STRESS CORROSION CRACK
BEHAVIOUR OF DISC SPRING
STACK IN 0.1 M CITRIC ACID
Figure 15 presents the lifespan of all variants in
0.1 M citric acid at 0.8h0 and 80 °C. The disc
spring variants composed of stainless steels withstand 2500 hours of the experiment without breaking. The same is true for the Dacromet- and Geomet-coated springs and for the oiled-only spring,
though with substantial surface corrosion attack.
The mechanically zinc-plated springs display the
shortest lifespan in citric acid of the six media
studied. As in 3 % NaCl solution, the springs
coated with Delta Tone and Delta Seal, the ones
with water-dilutable paint, and the chemically nickel-plated ones broke.
Spannungsrisskorrosionsversuch in 0,1 m Zitronensäure
1.4310 / C / D
1.4310 / C / D + K
1.4310 / B / D
1.4568 / C / S + D
1.4568 / C / S + D + K
1.4568 / C / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Lebensdauer [h] 0
2.500
2.500
2.500
2.500
2.500
2.500
68
68
2.500
2.500
526
380
262
2.500
891 h for Geo + Dacro!
891 h for Geo + Dacro!
500
839 h für geölt!
1.000
1.500
2.000
2.500
Figure 15: Comparison
of lifetime of all disc
spring variants in the
stress corrosion crack
experiment in 0.1 M
citric acid (0.8h0 and
80 °C)
23
1.4568 / C / S + D + K
1.4568 / C / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable lacquer
Figure 16:
Visual manifestation
of corrosion on the
disc springs in the
stress corrosion crack
experiment in 0.1 M
citric acid (0.8h0 and
80 °C)
Oiled
24
Figure 16 shows the different disc spring variants
after the experiment with 0.1 M citric acid. There
are no visual manifestations of corrosion to be
seen on the disc springs composed of stainless
steels, with the exception of the Kolsterised disc
spring. Citric acid belongs to the media that generate a passive layer. It can be assumed from the
results of the electrochemical study (not reported
here) that the damaged passive layer on stainless
steels in citric acid is even repaired. The disc
springs composed of stainless steels are protected against corrosion by the passive layer formed in air and in acid by their high chrome content. On the Kolsterised spring, a black layer consisting of carbon is formed.
The coated springs behaved much worse than the
stainless steels in 0.1 M citric acid. None of the
variants had a lifespan over 1000 h. It is known
from the electrochemical study (not reported here)
and the immersion experiment that the Zn portion
of the coating reacts chemically with the citric
acid. This caused cracks in the surface of the mechanically zinc-plated springs, which, due to the
excellent adhesion of the coating, extended all the
way into the base material. Moreover, the effectiveness of chromate layers only persisted up to
temperatures of 60°C. Both factors led to a drop
in lifespan of the mechanically zinc-plated disc
springs and thus to the shortest endurance time
(68 h).
In acid solutions of organic materials, zinc is not
resistant. The Dacromet and Geomet coatings dissolved in the citric acid. After some time, the
springs were completely free of coating (Dacromet
coating after about 20 h, Geomet coating after
about 50 h). Thus, at the end of the experiment,
these two disc spring variants displayed surface
corrosion like that of the oiled-only spring in citric
acid.
The disc spring coated with Delta Tone and Delta
Seal behaves better in citric acid than the other
coating variants. The coating was only destroyed at
the edges; the organic covering Delta Seal layer
acted as insulation and resisted acids and alkalis.
But due to the poor distribution of the covering
layer on the edges, only a short lifespan was
achieved.
have already been mentioned for the immersion
experiment, caused gap corrosion under the
coating. This causes a rapid penetration of the
corrosion into the depths of the base material
with the result of a premature wall collapse.
Susceptibility to cracking generally increases with
higher bending tension and rising temperature. If
the bending tension in the disc spring is too low,
the disc spring stack breaks not due to the crack
formation, but rather through dissolving (see Figure 7). Disc springs that come into contact with citric acid should optimally be produced from stainless steels.
The stress crack corrosion experiments were declared to have ended if a spring in the stack broke
or the experiment had lasted over 2500 hours.
When breakage occurred as cause of failure, it
was determined that the position of the broken
disc spring in the stack was relatively evenly distributed.
At the end of the experiments, it was determined
that the lifespan until failure could not be fully represented by the stress corrosion crack behaviour
of the disc spring. Thus, for example, the disc
springs shown in Figure 7 are evaluated as »no
failure« and thus as »good«; likewise with the oiled
disc springs shown in Figure 16 with a thick corrosion layer. In practice, this corrosion layer is presumably completely destroyed, which allows the corrosion to proceed rapidly until the disc spring
completely dissolves. A further evaluation option is
to add the loss of mass as a second evaluation
criterion. Table 9 gives the loss of mass for disc
springs whose experiments are listed in Table 7.
The mechanically zinc-plated and yellow chromatised spring failed after 68 hours with a loss of
mass of 1.347 g, the oiled spring lost a mass of
18.992 g without breaking, since the oiled spring
corroded on the surface over 2500 hours in 0.1 M
citric acid. This shows that the loss of mass does
not readily reflect the corrosion behaviour either.
Nevertheless, Table 9 shows that the Dacrometcoated and Geomet-coated springs had a »no
coating« behaviour similar to the oiled spring after
their coatings had been dissolved in the citric
acid.
The permeable areas in the paint layer and nickel
layer of the water-dilutable painteded and chemically nickel-plated disc springs, respectively, which
25
Disc spring variants
Seawater
MgCl2
NaCl
NaOH
Acid
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
1,9394 g
2,6715 g
0,6649 g
0,0059 g
0,6709 g
-0,1139 g
0,5094 g
1,3517 g
0,5482 g
-0,2885 g
0,7984 g
0,1439 g
-0,2747 g
0,5140 g
0,1108 g
1,8057 g
-0,2192 g
-0,2210 g
1,0361 g
0,0835 g
0,1009 g
0,1984 g
0,5552 g
0,8286 g
1,3707 g
1,6221 g
0,9095 g
1,6671 g
0,0049 g
0,1427 g
2,2457 g
1,9917 g
1,3470 g
1,6088 g
19,0022 g
19,0970 g
4,0592 g
4,2412 g
3,8797 g
18,9917 g
Table 9:
Loss of mass for disc
spring stack (6 x 1
coating) in the stress
corrosion crack experiment at 0.8h0 and
80 °C
Table 10:
Behaviour of disc
spring in stress
corrosion crack
experiment (80 °C)
In conclusion, a combination of the visual manifestation of corrosion with the lifespan until failure
is used to evaluate the corrosion behaviour. The
stress corrosion crack behaviour of the disc
springs was evaluated and classified according to
the following criteria:
■ Good:
• • Over 2500 h without failure
at 0.8h0 load and
• • Low manifestation of corrosion (as dots)
■ Moderate:
• • Lifespan between 1200 h and 2500 h
at 0.8h0 load or
• • Clear visual manifestation of corrosion
(as blotches)
■ Poor:
• • Lifespan between 300 h and 1200 h
at 0.8h0 load or
• • The springs are covered by thin corrosion
products (surface)
■ Very poor:
• • Lifespan less than 300 h at 0.8h0 load or
• • The springs are covered by thick corrosion
products (surface)
The disc springs evaluated as »good« are usable
in appropriate circumstances without reservation;
the disc springs evaluated as »very poor«, by contrast, should never be used. The use of disc
springs with the evaluation »moderate« or »poor«
must be weighed carefully case by case.
✩ – good
★ – moderate
✪ – poor
● – very poor
Disc spring variants
1.4310/C/S + D
1.4310/C/S + D + K
1.4310/B/S + D
1.4568/C/S + D
1.4568/C/S + D + K
1.4568/C/S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
26
Seawater
MgCl2
NaCl
NaOH
Acid
✩
✩
✩
✩
✩
✪
✪
✩
✩
✩
✩
✩
★
★
★
✩
✩
✩
✩
✩
✩
✩
✩
★
✩
✩
✩
✩
✩
✩
✩
✩
✪
✪
✪
●
✪
●
✪
✪
✩
✩
✪
✪
✪
✪
★
●
●
●
★
★
✩
✩
★
✪
✪
●
✩
★
★
✪
●
●
●
●
✪
✪
●
●
Disc spring variants
1.4310/C/S + D
1.4310/C/S + D + K
1.4310/B/S + D
1.4568/C/S + D
1.4568/C/S + D + K
1.4568/C/S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Geölt
Seawater
MgCl2
NaCl
NaOH
Acid
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
★
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
★
★
★
✩
✩
★
✩
✩
✩
★
★
★
✩
✩
✩
✩
✩
✪
✪
✪
✪
✪
✪
●
✪
✪
✩
★
★
✪
Tables 10 and 11 summarise the evaluation of the
stress corrosion crack behaviour of the disc
springs in the five media at 40 °C and 80 °C.
Most disc springs behave better at 40 °C than at
80 °C due to the reduction of the corrosion rate.
Temperature
40 °C
80 °C
Temperature
80 °C
Stress
0,6 h0
0,8 h0
0,6 h0
0,8 h0
Stress
0,4 h0
0,6 h0
0,8 h0
CORROSION BEHAVIOUR
OF 4 X 2 STACKED
DISC SPRING COLUMNS
Some selected disc spring variants were also studied in the form of a 4 x 2 disc spring stack and
some results are shown in Tables 12 and 13.
The lifespan of the 4 x 2 disc spring stack was
substantially longer than for the 6 x 1 layering.
✪
●
●
✪
✪
✪
✪
✪
✪
Table 11:
Behaviour of disc
spring in stress corrosion crack experiment
(40 °C)
✩ – good
★ – moderate
✪ – poor
● – very poor
1.4310 / C / S + D
6 x 14 x 2layering
layering
>2500 h
>2500 h
>2500 h
>2500 h
596 h
766 h
356 h
766 h
1.4310 / C / S + D + K
6 x 14 x 2layering
layering
>2500 h
>2500 h
>2500 h
>2500 h
476 h
>2500 h
429 h
1033 h
Table 12:
Comparison of the
lifespan of disc
springs composed of
1.4310 for different
layering arrangements in 40 % MgCl2
solution
Yellow chromatised
6 x 14 x 2layering
layering
220 h
2056 h
45 h
284 h
68 h
142 h
Transparently chromatised
6 x 14 x 2layering
layering
95 h
2032 h
69 h
142 h
68 h
116 h
Table 13:
Comparison of lifespan
of mechanically zincplated disc springs for
different layering arrangements in 0.1 M
citric acid
27
Figure 17: 4 x 2 disc spring stack placed under
tension at 0.2h0 (20 % deflection)
Figure 18: Visual manifestation of corrosion
on the mechanically zinc-plated disc springs in
citric acid
Through the parallel layering of disc springs in the
4 x 2 disc spring stack, additional friction is generated between the conical surfaces, which delays
the propagation of the cracks. Moreover, for the
4 x 2 disc spring stack, unlike for the 6 x 1
layering, some surfaces are not in contact with
the corrosion medium (see Figures 17 and 18).
should optimally be produced from stainless
steels. Almost all spring variants behaved well in
sodium hydroxide and deionised water. In the
media that contain chloride, the disc springs composed of stainless steels failed due to the high
chloride concentration (40 % MgCl2 solution) and
some coated variants due to the chloride concentration in combination with a high oxygen supply
(seawater atmosphere). The chemically nickelplated and painted disc spring variants failed in
almost all media.
The positions of the broken disc springs in the
4 x 2 layering are distributed relatively evenly,
similarly to the 6 x 1 layering.
SUMMARY OF
THE STRESS CRACK CORROSION
EXPERIMENTS
The Dacromet-coated and Geomet-coated springs
have the best stress corrosion crack behaviour in
four of the studied media, but not in citric acid.
The coated disc springs are not suitable for use in
citric acid due to the lack of chemical resistance.
The mechanically zinc-plated disc springs behaved
somewhat worse due to the lack of a barrier effect. Disc springs that come into contact with acid
28
Table 14 shows the results from the immersion experiment compared to those from the stress corrosion crack experiment. The results from the stress
corrosion crack experiment are significantly worse
than from the immersion experiment, i.e., despite
different experimental conditions, initial hints of
stress crack corrosion behaviour can already be
drawn from the immersion experiment.
Tellerfedervarianten
40 % MgCl2
3 % NaCl
0,1n NaOH
0,1m Säure
Tauchversuch
SpRK
Tauchversuch
SpRK
Tauchversuch
SpRK
Tauchversuch
SpRK
★
★
★
★
✪
✪
✩
✩
✩
✩
★
✩
✩
✩
✩
✩
★
★
★
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
★
✩
✩
✩
✩
✩
✩
✩
★
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
✩
1.4310/C/S + D
1.4310/C/S + D + K
1.4310/B/S + D
1.4568/C/S + D
1.4568/C/S + D + K
1.4568/C/S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
★
✪
●
●
●
●
✩
✩
✩
✩
✩
★
★
✩
✩
★
✪
✪
✪
●
✩
●
✪
✪
★
✩
✩
★
✪
✩
●
✪
✪
✪
●
✪
✩
★
★
●
●
●
●
✪
✪
✪
●
✪
●
●
●
●
✪
✪
●
●
Table 14:
Comparison of the
results from the immersion experiment
and the stress corrosion crack experiment
✩ – good
★ – moderate
VDA ALTERNATION TEST
WITH MECHANICAL LOAD
RESULTS OF
THE VDA ALTERNATION TEST
WITH MECHANICAL STRESS
AND DISCUSSION
✪ – poor
● – very poor
EXECUTION
VDA alternation tests with mechanical stress represent a combination of stress corrosion crack
experiments and the VDA alternation test (VDA
test sheet 612-415). A disc spring stack under
tension is exposed to the standardised VDA alternation test conditions [4].
The experimental conditions for the VDA alternation test with mechanical stress are similar to those
for the stress crack corrosion experiments in seawater atmosphere. The samples are placed in an
atmosphere that intermittently contains chloride.
The samples are exposed to a great deal of oxygen.
Figure 19:
Steps for the VDA
alternation test with
mechanical stress
29
Since the experimental temperature for the VDA
alternation test is significantly lower than 80 °C,
the lifespan in the VDA alternation test is longer
than for the stress corrosion crack experiments in
seawater atmosphere at 80 °C, but shorter than
at 40 °C. All results are shown in Table 15. For
stainless steels, only the Kolsterised variant failed.
As in the VDA alternation test, the yellow chromatised springs behave worse than the transparent
chromatised springs. Even the oiled and chemically nickel-plated variants failed here, but with decreasing preload the lifespan increases significantly. Permeable sites in the coating lead to
point corrosion at the edges of the chemically
nickel-plated and painted disc springs.
Figure 20 shows the visual manifestation of corrosion on the disc springs after the VDA alternation
test. The oiled, the Delta Tone/Delta Seal-coated,
the mechanically zinc-plated, and the Kolsterised
disc springs are covered by a thick corrosion product layer. No clear corrosion was visible on the
Dacromet-coated or Geomet-coated disc springs.
Most disc spring variants were covered by a thin
rust-brown layer. Permeable sites in the coating
lead to point corrosion at the edges of the waterdilutable painted and chemically nickel-plated
disc springs.
Disc spring variants
Table 15:
Results of all VDA
alternation tests with
mechanical load
30
1.4310/C/S + D
1.4310/C/S + D + K
1.4310/B/S + D
1.4568/C/S + D
1.4568/C/S + D + K
1.4568/C/S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
0,2 h0
SUMMARY OF
VDA ALTERNATION TEST
WITH MECHANICAL LOAD
In the VDA alternation test, the disc spring variants
composed of 1.4310 showed a very high resistance to corrosion due to the high proportions of
Cr and Ni in the material. The corrosion resistance
of the disc spring variants composed of 1.4568 is
somewhat worse, especially the Kolsterised variant, due to the high carbon content on the surface introduced by Kolsterising. In the VDA alternation test, a full Simonkolleite layer was not formed
on the zinc coating. After the zinc coating was
worn away, the base material corroded heavily. The
Dacromet coating and the Geomet coating offered
sufficient protection to the low alloyed springs. The
chemical nickel plating and the water-dilutable
paint are evaluated as poor due to the permeable
sites in the coatings. The oiled variant behaves the
worst, since the base material is only low alloyed
and the oil only offers short-term protection.
0,4 h0
0,6 h0
>2500 h
>2500 h
1636 h
0,8 h0
>2500 h
>2500 h
>2500 h
>2500 h
>2500 h
1228 h
2043 h
>2500 h
>2500 h
>2500 h
>2500 h
841 h
>2500 h
1924 h
1.4310 / C / S + D
1.4568 / C / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Dacromet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
Figure 20:
Visual manifestation
of corrosion on the
disc springs after the
VDA alternation test
with mechanical load
With water-dilutable paint
Oiled
31
FATIGUE CRACK
CORROSION EXPERIMENTS
which is also electrically insulating. Initial failures
of the test equipment were able to be fixed using
a suitably constructive structure.
Similar to the stress corrosion crack experiment,
the disc springs are subjected to a complex stress
(mechanical and corrosive), but in contrast to the
stress corrosion crack experiment, an alternating
mechanical load (cyclic) is applied here.
The test setup for the fatigue crack corrosion experiments (Figures 21 and 22) consists of three
modules: servohydraulic test machine (maximum
force 63 kN, maximum stroke 100 mm), corrosion
chamber, and storage vessel with pump.
EXECUTION OF THE FATIGUE
CRACK CORROSION EXPERIMENTS
RESULTS OF THE FATIGUE
CRACK CORROSION EXPERIMENTS
For these experiments, special requirements were
placed on the execution. The 6 x 1 disc spring
stacks had to be used with an internal guide without grease and thus continuous irrigation with the
corrosion medium took place. No complete immersion of the stack in the corrosion medium occurred
during the experiment in order to avoid creating a
build-up of hydrostatic pressure in the inner region
of the stack, which could cause additional forces
on the disc springs.
Table 16 summarises the cycles until failure for
the disc springs in the fatigue crack corrosion experiment (cycle width 0.2h0 – 0.8h0) in the different media. Experiments in seawater atmosphere
(hard to achieve) and in 40 % MgCl2 solution
(highly aggressive) were not performed. A 3 %
NaCl solution represented the chloride solutions.
The tools for test arrangement for using the columns were produced from a ceramic material of
sufficient hardness, corrosion, and wear resistance
The disc spring variants composed of stainless
steels had the longest lifespan in 0.1 N sodium
hydroxide, the shortest in deionised water. The
lifespan decreased from 0.1 N sodium hydroxide
through 0.1 M citric acid and 3 % NaCl solution to
deionised water. The deionised water as a medium
Test machine
Corrosion chamber
Pump
Storage chamber
Figure 21 Test setup for fatigue
crack corrosion experiment
32
Figure 22:
Corrosion chamber with load device
could not electrochemically or chemically destroy
the passive layer formed in the air, meaning that
the stainless steels are resistant to corrosion in
deionised water. Under fatigue condition , however,
local formation of slip bands was able to destroy
the passive layer. The regions thereby exposed
could not be repassivated and became electrochemically active sites (local anodes) on the surface. A local anode created in this way could
achieve stability due to the unfavourable surface
relation of anode to cathode surface, the resulting
local high material conversion, and the decrease
in pH at the bottom of the hole, and lead to the
formation of a notch.
In 3 % NaCl solution, the results from the immersion experiment showed that the chloride ions broke
through the air-passivated layer. Multiple active
centres were created on the disc spring surfaces,
which made the surface ratio of anode surface to
cathode surface more favourable, hence the material conversion lower and the lifespan longer.
The coated disc spring variants had the longest
lifespan in 0.1 N sodium hydroxide, the shortest in
0.1 M citric acid. As with the disc spring variants
of stainless steels, the longest lifespan could be
traced to the hydroxide layer formation in 0.1 N
sodium hydroxide, which protects the disc springs
against corrosion. Zinc layers react with citric acid,
which had already been seen in the immersion
experiment and the stress corrosion crack experiment. This chemical reaction led to crack formation in the layer that then continued into the base
material. In addition, the low alloyed material
1.8159 was very susceptible to corrosion in
0.1 M citric acid due to the low pH value (2.09).
The higher conductivity of the 3 % NaCl solution
(about 42.4 mS/cm) induced a shorter lifespan
for some coated disc spring variants than in deionised water (about 0.4 mS/cm). For others, the
opposite was the case. The Simonkolleite layer
formed in the chloride-containing solution continuously repaired the damaged sites and failure
was delayed.
In 0.1 N sodium hydroxide and 0.1 M citric acid,
the active centres were covered again due to repassivation of the passive layer; the capacity and
rate of repassivation in sodium hydroxide is better
than in citric acid. This is confirmed in electrochemical studies and in the literature.
Disc spring variants
1.4310/C/S + D
1.4310/C/S + D + K
1.4568/C/S + D
1.4568/C/S + D + K
1.4568/C/S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Deionised
water
14.171
18.255
12.924
20.480
11.339
26.839
7.841
5.676
5.428
24.975
7.083
22.138
13.956
NaCl
NaOH
Zitronensäure
17.952
20.300
17.207
24.823
22.199
25.510
11.323
4.944
6.159
10.355
6.461
13.469
5.493
37.767
38.033
32.747
34.555
32.533
26.477
14.509
6.033
4.517
10.127
12.058
9.902
19.606
22.280
25.389
19.520
20.090
30.883
14.058
4.318
4.849
4.031
5.563
6.414
4.195
5.178
Table 16:
Lifespan (fatigue
cycles until failure, N)
of the disc springs in
6 x 1 stack for the fatigue crack corrosion
experiments (cycle
width 0.2h0 to 0.8h0)
33
Disc spring variants
Table 17:
Lifespan (fatigue cycles until failure, N)
of the disc springs in
6 x 1 stack for the fatigue crack corrosion
experiments (cycle
width 0,2h0 to 0,6h0).
1.4310/C/S + D
1.4310/C/S + D + K
1.4568/C/S + D
1.4568/C/S + D + K
1.4568/C/S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone + Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Deionised
water
19.552
33.236
12.357
21.845
32.933
103.618
153.506
129.507
141.642
167.443
47.429
94.033
106.702
In Table 17, the results of the fatigue crack corrosion experiments are listed with a cycle width
0.2h0 r
R 0.6h0. By reducing the amplitude of the
travel, the lifespan of the coated disc spring variants was significantly lengthened, but those of the
stainless steel variants only to a small extent.
Figure 23 summarises the lifespan values of all
disc spring variants for the fatigue crack corrosion
R 0.6h0) in
experiment (cycle amplitude 0.2h0 r
deionised water.
NaCl
21.858
40.005
17.383
27.974
34.000
292.537
295.742
46.388
59.555
240.707
27.854
91.741
32.806
NaOH
51.965
1.702.463
1.443.281
Zitronensäure
30.037
47.338
34.692
41.433
40.250
73.386
49.507
28.192
24.128
22.578
19.208
15.703
28.078
The coated disc springs had a significantly longer
lifespan than the stainless steel disc spring variants. The passive layer (oxide film) on the stainless steels during fatigue was broken through in a
highly localised manner by a slip band penetrating
outwards. This region now formed a small, highly
active local anode. The corrosion attack followed
the slip band into the material. In deionised water,
the materials 1.4310 and 1.4568 were not in a
position to regenerate the original passive layer at
the sites of corrosion. The notches created by the
material conversion that had already occurred
Vibration crack corrosion experiments in deionised water
Figure 23:
Comparison of lifetime
of all disc spring
variants in the stress
corrosion crack experiment in deionised
water (cycle amplitude
0,2h0 to 0,6h0).
34
1.4310 / S + D
1.4310/ S + D + K
1.4568 / S + D
1.4568 / S + D + K
1.4568 / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone / Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Fatigue cycles [N] 0
19.552
33.236
12.357
21.845
32.933
103.618
153.506
129.507
141.642
167.443
47.429
94.033
106.702
50.000
100.000
150.000
200.000
then led to a local increase in stress; the crack
progressed at an increasing rate until breakage
occurred [8]. No passive layer was formed on the
coatings in air. Thus, the coated disc springs did
not experience this problem in deionised water.
In addition, numerous places in the literature
show that the material 51CrV4 displays a better
resistance to fatigue than the materials 1.4310
and 1.4568 [9]. Shot-peening and Kolsterising
had a positive effect during the fatigue crack
corrosion experiment. The residual compressive
stress generated near the edge reduced the
mean stress of the fatigue load. The oiled spring
endured 106,702 cycles. Significantly longer or
shorter lifespans always indicated the involvement of the coating. Thus, for example, the chemically nickel-plated spring failed after 47,429
cycles due to the permeable site in the coating,
through which the base material corroded at an
accelerated rate.
In 3 % NaCl solution, the coated disc spring variants showed, on average, better fatigue crack
corrosion behaviour than the stainless steel disc
spring variants (Figure 24). The disc springs composed of 1.4310 had a longer lifespan due to
the better corrosion resistance than the disc
springs composed of 1.4568; in 3 % NaCl solution, cracks in the disc springs composed of
1.4310 and 1.4568 were initiated by chlorideinduced pitting corrosion, which can be viewed
as an operation-induced notch effect. Many
crack nuclei were created, which eventually expressed themselves as transcrystalline running
cracks. As a result, in this case, the break profile
can be clearly distinguished from a fatigue failure
in air.
No impermeable protective layer could form on
the oiled disc spring in air or in 3 % NaCl solution,
so fatigue crack corrosion occurred in the active
state. Corrosion grooves were created, which acted
as notches from which cracks formed due to the
component of mechanical stress. Due to the
worse corrosion resistance in 3 % NaCl solution
than in deionised water, the oiled spring had a
shorter lifespan. This applied in a similar way for
the chemically nickel-plated, the Dacrometcoated, and the Geomet-coated springs. The painted spring had almost the same lifespan as in the
deionised water, that is, the fatigue resistance of
the paint layer is the deciding factor for the fatigue
crack corrosion behaviour in deionised water and
in 3 % NaCl solution. The Dacromet and Geomet
coatings are brittle and only adhere to the base
material to a limited extent. After only a few cycles, the layers peeled away in flakes. Thus, the
Dacromet- and Geomet-coated disc springs behaved in all media like the oiled spring. The mechanically zinc-plated disc spring and the disc spring
coated with Delta Tone and Delta Seal endured a
higher number of cycles than in deionised water
through the formation of the Simonkolleite layer
on top of the zinc layer.
Vibration crack corrosion experiments in 3 % NaCl solution
1.4310 / S + D
1.4310/ S + D + K
1.4568 / S + D
1.4568 / S + D + K
1.4568 / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone / Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Fatigue cycles [N] 0
21.858
40.005
17.383
27.974
34.000
292.537
295.394
46.388
59.555
240.707
27.854
91.741
32.806
50.000 100.000
150.000 200.000 250.000 300.000
Figure 24:
Comparison of lifespans of all disc
spring variants in the
fatigue crack corrosion
experiment in 3 % NaCl
solution (cycle amplitude 0,2h0 to 0,6h0).
35
Fatigue crack corrosion experiments in 0.1 M citric acid
Figure 25:
Comparison of lifespan
values of all disc
spring variants in the
fatigue crack corrosion
experiment in 0.1 M
citric acid (cycle
amplitude 0,2h0 to
0,6h0).
1.4310 / S + D
1.4310/ S + D + K
1.4568 / S + D
1.4568 / S + D + K
1.4568 / S + D + Kolsterised
Mechanically zinc-plated + yellow chromatised
Mechanically zinc-plated + transparently chromatised
Dacromet-coated
Geomet-coated
Delta Tone / Delta Seal
Chemically nickel-plated
With water-dilutable paint
Oiled
Fatigue cycles [N] 0
30.037
47.338
34.692
41.433
40.250
73.386
49.507
28.192
24.128
22.578
19.208
15.703
28.078
20.000
In 0.1 M citric acid (Figure 25), the disc spring variants in stainless steel have on average a longer
lifespan than the coated disc spring variants. The
chemical reaction between citric acid and zinc,
present in most coatings, dissolved the coatings.
The water-dilutable paint layer was not chemically
resistant in 0.1 M citric acid, either. These factors
led to earlier failure of the coated disc springs in
0.1 M citric acid than in deionised water or 3 %
NaCl solution. From the results of the electrochemical studies (not reported here), it was known
that citric acid is a medium that produces a passive layer for the passivatable chrome-nickel
steels. Thus, the stainless steel disc spring variants
here experience fatigue crack corrosion in the
passive state.Cyclic loading creates surface regions free of a passive layer via slip bands that penetrate outwards. These regions now formed small,
Figure 26:
Visual manifestations
of corrosion on nonshot-peened disc
springs composed of
1.4310 after the fatigue crack corrosion
experiment in 3 %
NaCl solution
36
40.000
60.000
80.000
highly active local anodes. The corrosion attack
followed the slip band into the material. In citric
acid, the chrome-nickel steels were in a position
to regenerate the original passive layer at the sites
of corrosion. The crack then predominantly came
to a halt. However, the notch created by the material conversion that had already occurred led after
some time, with increase in notch formation, to a
locally elevated tensile stress, which cracked the
just-formed passive layer again. The attack then
always continued into the depths of the material.
The crack propagated at an increasing rate [8].
The chrome-nickel steels had a better repassivation ability in 0.1 N NaOH than in 0.1 M citric acid,
as shown by a correspondingly longer lifespan.
In contrast to the even distribution of breaks over
the disc spring stacks in the stress corrosion crack
250.000
200.000
150.000
230.305
19.606
5.178
5.493
13.956
76.299
32.747
19.520
17.207
12.924
32.156
37.767
22.280
17.952
50.000
14.171
VFatigue cycles [N]
100.000
0
1.4310 / S + D
Disc spring variants
experiment, in the fatigue crack corrosion experiment it was predominantly the second spring in
the 6 x 1 stack that failed, since due to the experimental arrangement, the second spring corroded
most heavily on the upper and lower sides (Figure
26).
In the context of the research project AVIF A115
»Fatigue resistance studies of individual disc
springs and disc spring stacks of freely variable
layering arrangements« and the continuation project AVIF A155 »Supplementary studies of disc
springs«, the fatigue resistance properties of disc
springs in air were studied. In Figure 27, the results from both research projects are compared
with the results from this research project.
The negative influences and effects of corrosion
on component resistance to cyclic operational
load have been shown by experience to be more
severe as the strength of the construction material
increases. Conventional, low alloyed spring steels
offer no protection of their own against corrosion.
Thus, for springs which must in principle be comprised of high-strength materials that are relatively
sensitive to notching in order to fulfill their function, sufficient protection against these influences
is necessary. Damage to the surface of the spring
through chemical or electrochemical attack is to
be avoided to the greatest extent possible, since
even the smallest corrosive pits can lead to breakage.
1.4568 / S + D
1.8159 / Oiled
Figure 27:
Comparison of lifespan
values of all disc
spring variants in
corrosion media and
in air (cycle amplitude
0,2h0 to 0,8h0).
Deionised water
NaCl
Acid
NaOH
Air
SUMMARY OF THE FATIGUE CRACK
CORROSION EXPERIMENTS
In deionised water, the disc springs composed of
stainless steels had the shortest lifespan (fatigue
cycles endured). With relation to medium, their lifespan was in the following order: sodium hydroxide
> citric acid > NaCl > deionised water. The shotpeening had a positive effect, independent of material, on the lifespan of the spring, especially at
small amplitudes of cycle. The coated disc springs
composed of 51CrV4 generally had better fatigue
resistance properties than the disc springs of
stainless steels, except in citric acid. Of the coated
disc springs, the mechanically zinc-plated disc
springs behaved the best; the Dacromet and Geomet coatings are brittle and not suitable for cyclically stressed disc springs.
37
38
BIBLIOGRAPHY
[1] van Eijnsbergen J.F.H.
Duplex Systems: hot-dip galvanizing plus painting.
Elsevier Science BV. 1994
[2] MUBEA – Disc Springs (Manual) [Tellerfedern (Handbuch)];
Firma Muhr und Bender, Daaden 1992
[3] Falbe, J. und Regitz, M.
Römpp Chemistry Lexicon, 9th Edition [Chemie Lexikon, 9. Auflage];
Georg Thieme Verlag Stuttgart, New York 1995
[4] VDA Test Sheet [Prüfblatt] 621-415, »Testing corrosion protection
of vehicle paint coats under cyclically alternating stress« [Prüfung
des Korrosionsschutzes von Kraftfahrzeuglackierungen bei zyklisch
wechselnder Beanspruchung]; Verband der Automobilindustrie e.V.,
Frankfurt, February 1982
[5] Spähn, H.
Paper on early recognition of damage due to cavitation erosion
[Beitrag zur Schädigungsfrüherkennung bei Kavitationserosion];
VDI-Verlag, Düsseldorf 1993
[6] Li, Y.: Corrosion behaviour of disc springs and disc spring stacks under
complex stress [Korrosionsverhalten von Tellerfedern und Tellerfedersäulen
unter Komplexbeanspruchung].
Dissertation TU Darmstadt, Shaker Verlag, Aachen 2007
[7] Shot-peening in subcontracting, 2nd edition [Kugelstrahlen
im Lohnauftrag, 2. Ausgabe]; Kiefer GmbH, Ettlingen 2003
[8] Grote, K.-H. und Feldhusen, J.
Pocket book for machine construction [Taschenbuch
für den Maschinenbau]; Dubbel Springer Verlag, Berlin 2004
[9] Teller, Cord;
Fatigue resistance of disc spring stacks depending on layering
arrangement, material, and preparation; reports from the material
technology [Schwingfestigkeit von Tellerfedersäulen in Abhängigkeit
von Schichtung, Werkstoff und Fertigungszustand; Berichte
aus der Werkstofftechnik]; Shaker Verlag, Aachen 2002
39
Telefon: 0 71 82/ 12-0
Telefax: 0 71 82/ 12-315
E-Mail: info@christianbauer.com
Internet: www.christianbauer.com
England
Bauer Springs Ltd.
Eagle Road
North Moons Moat Ind. Estate
GB-Redditch Worcs. B98 9HF
Telefon: + 44/ 15 27-594 900
Telefax: + 44/ 15 27-594 909
E-Mail: sales@bauersprings.co.uk
Internet: www.bauersprings.co.uk
USA
Bauer Springs Inc.
509 Parkway View Drive
Parkway West Ind. Park
USA-Pittsburgh, PA. 15205
Telefon: +1/ 412-787-79 30
Telefax: +1/ 412-787-38 82
E-Mail: info@bauersprings.com
Internet: www.bauersprings.com
1002 markohaag.de
Christian Bauer
GmbH + Co. KG
Schorndorfer Straße 49
D-73642 Welzheim
CORROSION BEHAVIOUR
OF DISC SPRINGS